RDP sessions broken

[AndiK]

We have 5 Windows Server 2003 Enterprise Edition in a wlbs cluster with
clustered session directory serving about 250 users with thin clients. Each
Server has to network interfaces, one inside the wlbs cluster, the other
intended for backend traffic.
Most of the thin clients are Wyse Thin OS devices (S10 and Winterm 1200LE)
with "legacy" rdp clients. Clients and Servers are on our inhouse LAN (Cisco
switches 1GBit/s Backbone, 100MBit/s at the client side).
Some users with that Wyse Thin Clients experience suddenly broken rdp
connections, sometimes only once or twice a day, on other days 3 or four
times in 5 minutes. I suppose that some users, even though affected, don't
know about the problem due to the reconnect feature of the tc devices and the
session directory.
Devices with microsoft rdp clients (5.2 and 6.0 on XPe Thin Clients and XP
Workstations) work without any losses of rdp connections.
I tried a lot of different workarounds and settings but cannot fix this
issue. Is there anyone outside who experiences similar problems or does
anyone have an idea how to fix this issue?

Best regards

Andi

 

[AndiK]

RE: RDP sessions broken

Not the solution (i hope that our LAN Division will provide one , but an
effective workaround in my case was setting TcpMaxDataRetransmissions to
10(dez).

<Quote>
In the registry at HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters,
create or edit the DWORD value of TcpMaxDataRetransmissions. By default it is
set to 5, but I would recommend doubling that value, to 10. The value of
TcpMaxDataRetransmissions is the number of times TCP retransmits an
unacknowledged data segment on an existing connection. TCP retransmits data
segments until they are acknowledged or until this value expires. Basically,
when a client doesn't respond to a packet from the terminal server, the
server will attempt to retransmit the packet up to TcpMaxDataRetransmissions
number of times. By increasing this value, you are giving the client more
time to respond to the server, which will help improve flaky connections or
connections with high latency or higher than normal packet loss.
</Quote>

"AndiK" wrote:

> We have 5 Windows Server 2003 Enterprise Edition in a wlbs cluster with
> clustered session directory serving about 250 users with thin clients. Each
> Server has to network interfaces, one inside the wlbs cluster, the other
> intended for backend traffic.
> Most of the thin clients are Wyse Thin OS devices (S10 and Winterm 1200LE)
> with "legacy" rdp clients. Clients and Servers are on our inhouse LAN (Cisco
> switches 1GBit/s Backbone, 100MBit/s at the client side).
> Some users with that Wyse Thin Clients experience suddenly broken rdp
> connections, sometimes only once or twice a day, on other days 3 or four
> times in 5 minutes. I suppose that some users, even though affected, don't
> know about the problem due to the reconnect feature of the tc devices and the
> session directory.
> Devices with microsoft rdp clients (5.2 and 6.0 on XPe Thin Clients and XP
> Workstations) work without any losses of rdp connections.
> I tried a lot of different workarounds and settings but cannot fix this
> issue. Is there anyone outside who experiences similar problems or does
> anyone have an idea how to fix this issue?
>
> Best regards
>
> Andi